The work in my laboratory has focused on enhancements of gene transfer vectors for cancer gene therapy. Viral gene transfer vectors are invaluable tools for the study of gene expression and for clinical gene therapy applications. Moloney murine leukemia virus (MoMLV)-based vectors have been extensively utilized for this purpose owing to their ability to stably transfer the gene of interest to target cells. However, a major limitation to the use of these vectors is the relative inability to develop stable vector producer cells (VPC's) capable of generating retroviral vectors expressing toxic genes. Previously reported strategies to overcome this problem utilizing tissue specific or inducible promoters exhibit some degree of "leakiness." Over the last year, we generated a novel MoMLV-based retroviral vector encoding a reverse transcription-activated transgene (RETRACT) that precludes expression of the gene in the vector producer cells. The arrangement of the RETRACT vector system allows expression of the transgene only in target cells. A prototypical RETRACT vector was generated by cloning the cDNA of interest upstream of the viral 3' LTR or in the U3 region of the 3' LTR in reverse orientation relative to the viral transcriptional unit without a promoter to drive expression of the transgene in the VPC's. A strong exogenous promoter was cloned in reverse orientation at the R-U5 border of the viral 5' LTR. On transduction of target cells, the natural reverse transcription of the vector RNA copies the 5'LTR containing the promoter to the 3' LTR, where it then drives expression of the reverse-oriented transgene. We tested this system using green fluorescent protein (GFP) as a marker gene and the SV40 promoter. The vector was also designed with the E. coli cytosine deaminase (CD) gene in the forward orientation under the control of the retroviral 5' LTR allowing for its constitutive expression as a negative selection marker. The PG13 derived RETRACT VPC's gave viral titers of 1-5 x 104 c.f.u./mL. By flow cytometry, GFP expression was not detectable in the PG13-RETRACT VPC's, however, fluorescence was seen in TE671 target cells transduced with viral supernatants from the VPC's. The fluorescence intensity was improved by the generation of a double copy vector (RETRACTdc). Southern blot and PCR analysis of genomic DNA from the VPC's and the TE671 target cells demonstrated the approximation of the SV40 promoter to the GFP transgene only in the latter. Northern analysis of the producer and the target cells using a GFP-specific probe demonstrated an appropriately sized GFP transcript only in the TE671 target cells that would be expected when the GFP was driven by the copied SV40 promoter. We are currently working on generating stable RETRACT VPC's expressing the diphtheria toxin A-chain to test the feasibility of this approach. The ultimate goal is to develop packaging cells lines producing retroviral vectors expressing cytotoxic genes that would continuously produce vector and yet be unaffected by the gene or their own retrovirus. These producer cells could be inoculated into a tissue or organ (i.e. a tumor or perhaps a destructive rheumatoid pannus) and transduce the target tissue with the RETRACT vector over a prolonged period of time. Once the desired effect is achieved, the VPC's could be removed by treatment with 5-fluorocytosine, which is activated to a toxic metabolite (5-fluorouracil) by the CD enzyme.